The southern Strait of Georgia, British Columbia, often contains packets of large, near‐surface internal waves. Wave crests at the leading edge of the packet, spaced a few hundred meters apart, can have a longitudinal extent of more than 10 km. It has long been assumed that these waves are generated by tidal flow through narrow passages and channels at the Strait's southern boundaries, but no actual link has ever been made between these waves and a specific passage or generation mechanism. Here we identify the location and extent of a number of these large packets at specific times using mosaics of photogrammetrically rectified oblique air photos. Wave speeds are determined by analyzing a time sequence of images, with water column measurements used to subtract effects of tidal advection. The location and extent of these internal waves are then compared with the predicted location and extent of hypothetical waves generated in different passages, at different stages of the tide, which are then propagated through a predicted time‐varying barotropic flow field. It is found that the observed waves are most likely generated near or after the time of the peak flood tide, or peak inflow into the Strait. They are therefore inconsistent with generation mechanisms involving the release and upstream propagation of waves by the relaxation of an ebb tide. Instead they are probably associated with the nonlinear adjustment of conditions at the edge of an inflowing injection of relatively weakly stratified water.

The observations reported here are based on time series of in situ observation data in Laoshan Bay off the Qingdao coast. A chain of thermistors (T‐chain) at a fixed location recorded a sequence of elevation internal waves followed by depression internal waves passing by over an elapsed time of about 1 h. This observed polarity conversion at a fixed location is caused by the vertical stratification variation mainly induced by the rising tide, which is believed to be the first reported observation of this kind. The process of an elevation internal wave train converting to a depression wave train is simulated using the variable‐coefficient extended Korteweg‐de Vries (veKdV) equation, which also provides a further comparison between theory and the reported observations.

No abstract

There are now several observations of internal solitary waves passing through a critical point where the coefficient of the quadratic nonlinear term in the variable coefficient Korteweg–de Vries equation changes sign, typically from negative to positive as the wave propagates shoreward. This causes a solitary wave of depression to transform into a train of solitary waves of elevation riding on a negative pedestal. However, recently a polarity change of a different kind was observed in Laoshan Bay, China, where a periodic wave train of elevation waves converted to a periodic wave train of depression waves as the thermocline rose on a rising tide. This paper describes the application of a newly developed theory for this phenomenon. The theory is based on the variable coefficient Korteweg–de Vries equation for the case when the coefficient of the quadratic nonlinear term undergoes a change of sign and predicts that a periodic wave train will pass through this critical point as a linear wave, where a phase change occurs that induces a change in the polarity of the wave, as observed. A two-layer model of the density stratification and background current shear is developed to make the theoretical predictions specific and quantitative. Some numerical simulations of the variable coefficient Korteweg–de Vries equation, and also the extended variable coefficient Korteweg–de Vries equation, are reported that confirm the theoretical predictions and are in good agreement with the observations.

The linkage between physical and biological processes is studied by applying a one-dimensional physical-biological coupled model to the Sargasso Sea. The physical model is the Princeton Ocean Model and the biological model is a five-component system including phytoplankton, zooplankton, nitrate, ammonium, and detritus. The coupling between the physical and biological model is accomplished through vertical mixing which is parameterized by the level 2.5 Mellor and Yamada turbulence closure scheme. The coupled model investigates the annual cycle of ecosystem production and the response to external forcing, such as heat flux, wind stress, and surface salinity, and the relative importance of physical processes in affecting the ecosystem. Sensitivity experiments are also carried out, which provide information on how the model bio-chemical parameters affect the biological system. The computed seasonal cycles compare reasonably well with the observations of the Bermuda Atlantic Time-series Study (BATS). The spring bloom of phytoplankton occurs in March and April, right after the weakening of the winter mixing and before the establishment of the summer stratification. The bloom of zooplankton occurs about two weeks after the bloom of phytoplankton. The sensitivity experiments show that zooplankton is more sensitive to the variations of biochemical parameters than phytoplankton.